                  
                  
                  
                  
                  
                  
                  
                  
                  
                  
                  
                  
                  
                  
                  
                         Political Advertising in 2000











            A Dataset Compiled by the Wisconsin Advertising Project




















                          Final Release, December 2002
                                        
                                        
                                        
                                        
                                        
                                        
                                        

 Wisconsin Advertising Project, University of Wisconsin-Madison, Department of
     Political Science, 110 North Hall, 1050 Bascom Mall, Madison WI 53706,
                            wiscads@polisci.wisc.edu


BIBLIOGRAPHIC CITATION

Publications based on this data collection should
acknowledge this source by means of bibliographic
citation. The bibliographic citation for this data 
collection is:

Goldstein, Kenneth, Michael Franz, and Travis Ridout. 2002.
"Political Advertising in 2000." Combined File [dataset].
Final release. Madison, WI: The Department of Political 
Science at The University of Wisconsin-Madison and
the The Brennan Center for Justice at New York University.

Users of these data should also include the following 
acknowledgement:

These materials are based on work supported by the Pew
Charitable Trusts under a grant to the Brennan Center for
Justice at New York University and a subsequent sub-contract
to The Department of Political Science at The University of
Wisconsin-Madison.


DATA COLLECTION DESCRIPTION

The data were purchased from Campaign Media Analysis Group
(CMAG), a company based in Alexandria, Virginia.  CMAG uses
information gathered by Competitive Media Reporting (CMR)
and specifically, CMR's "Ad Detector" product.   CMR is one
of the country's leading providers of marketing
communication and advertising expenditure information to
advertising agencies, advertisers, broadcasters and
publishers.  CMR's "Ad Detector" technology was originally
developed for large corporations to track competitors'
advertisements and to confirm that their own advertisements
were being aired.  In 2000, the system monitored the
satellite transmissions of the national networks (ABC, CBS,
NBC, and Fox) as well as 25 national cable networks (such as
CNN, ESPN, and TBS).  In addition, the system monitored
advertising in the country's top 75 markets. (Although there
are over 200 media markets in the United States, over 80
percent of the population lives in the top 75 markets.)  The
system's software recognizes the electronic seams between
programming and advertising.  When the system does not
recognize the unique digital code of a particular commercial
spot, the storyboard (the full audio and every four seconds
of video) is captured and downloaded to the firm's
headquarters. 

After reviewing the storyboards, analysts at CMR then place
the advertisements into particular categories -- by product
for commercial clients, by candidate or sponsor for
political clients -- and tag them with unique digital
fingerprints.  Thereafter, the system automatically
recognizes and logs that particular commercial wherever and
whenever it airs.  All political spots are flagged and
immediately transferred to CMAG - which markets the data to
political clients (candidates, parties, and interest
groups). 

The compilation of political advertising data in the 2000
elections took place in two distinct stages:  the
pre-election period (referring to coding and merging done up
through December 2000) and the post-election period
(referring to data cleaning, missing data issues, and
additional coding and merging beginning in January 2001).

Two University of Wisconsin-Madison graduate students, along
with Professor Ken Goldstein, helped manage the receipt and
compilation of two types of information during the summer
and fall of 2000: storyboard (content) and ad frequency
data.  The frequency data were in files containing
information on the timing, date, market location, station,
show, and estimated cost of each political spot airing. 
They did not contain information on the content of the
advertisements.  CMAG sent by e-mail new frequency data
every two weeks during the summer, and weekly during the
fall of 2000.  On about the same schedule, CMAG regularly
mailed one-page storyboards, which are captured visual clips
of the ad taken every 3-4 seconds along with the
accompanying transcript.  CMAG would mail a storyboard for
each unique creative.

Both sets of data were combined into a larger database.  
First, the frequency data were placed into a large
spreadsheet.  Second, the graduate students coded each
storyboard using an online coding sheet, accessed through
the CMAG website.  Approximately 35 questions were asked. 
(The coding questions were first developed by in spring of
1999 by a committee consisting of Josh Rosenkranz of the
Brennan Center at New York University,  Jon Krasno of the
Brennan Center at New York University, Tom Mann of the
Brookings Institute, David Magleby of Brigham Young
University, and Darrel West of Brown University.  The
original coding protocol was developed for 1998. Changes
were made to reflect new issues.)  Coding at the website was
automatically translated into a separate database, which
would periodically be downloaded from CMAG and linked to the
database containing the frequency data.

As the volume of storyboards increased in September and
October, a team of six undergraduates began to assist with
the coding.  Each undergraduate was given a set of
storyboards, and the graduate students logged them on to the
CMAG coding website.  The undergraduates were given basic
instruction on how to operate the coding website, but were
instructed to complete the coding as if they were answering
questions on a survey instrument.  In other words, no
additional instructions were given on particular questions
or on the goals of the project as a whole. 

In early January of 2001, frequency information and data
from the coded storyboards (content information) were
merged. Additional quality control was the primary concern
of the post-election period.

An initial comparison of frequency data with coded
storyboards produced three forms of "missing information:"
1) ad storyboards without accompanying frequency data; 2)
frequency data without accompanying ad storyboards; and 3)
storyboards that remained uncoded. 

The project requested from CMAG frequency data for the
storyboards that lacked it and obtained such information for
almost all storyboards.  The missing frequency data
generally came from down-ballot races.  The project was also
able to obtain storyboards for all but eight ads listed in
the frequency data.  The graduate students coded storyboards
that had slipped through the cracks, either because of an
occasional glitch in the CMAG computer system that would not
update coding when it was entered or because of an
undergraduate coder who lost some storyboards.  Some ads
concerning low-level races (such as a state senate race)
were never coded, though frequency data were included in the
database.  Missing data codes were put in the data and noted
in the codebook.

The project then dealt with "cookie cutter" ads, ads run by
interest groups in several congressional districts that are
identical except for the name of the candidate.  Because the
CMAG tracking system cannot identify these slightly
different ads as unique, all cookie cutter airings are
initially assigned to the same candidate.  This yields a
situation in which, for example, CMAG reports that an ad
running against Rep. Anne Northup of Kentucky aired in
Kentucky as well as Minnesota.  In truth, however, the ad
running in Minnesota used not Northup's name but that of a
different candidate.  Thus, the project had to assign the
ads running in Minnesota to a Minnesota race, instead of the
Kentucky race.  This was done by comparing maps of media
markets and congressional districts.  If an ad ran in the
Minneapolis-St. Paul media market, the project identified
all congressional districts which the media market covered
and then using information on competitive races and
conversations with Washington based consultants and party
officials, the project determined the actual targeted
district.  

In addition, the project corrected an additional problem
associated with cookie cutter advertising.  After assigning
a cookie cutter ad to a new district, associated contextual
information (such as the incumbency status of the favored
candidate) had to be changed.

ESTIMATING AD COST

CMAG estimates the cost of an ad by using market information
on the average cost of an ad during the time during which
that ad aired.  This estimate is likely an underestimate of
the true expenditure, especially in the case of interest
group and party advertising.

DATA RELIABILITY

For further information about the reliability of the data
collection and its coding, please see the following
manuscript:  Ridout, Travis N., Michael Franz, Kenneth
Goldstein and Paul Freedman. 2002. "Measuring
Exposure to Campaign Advertising."  It is available at
www.polisi.wisc.edu/~tvadvertising/reliability.pdf.



VARIABLE NAMES

Custitle:  Unique name given to each ad by CMAG

Len:  Length of ad in seconds

Daypart:
	Early morning:  6-10 a.m.
	Daytime:  10 a.m. -  4 p.m.
	Early Fringe:  4 p.m. - 7:30 p.m.
	Early News:  News in Early Fringe
	Prime Access:  7:30 p.m. - 8 p.m.
	Prime Time:  8 p.m. - 11 p.m.
	Late news:  10 or 11 p.m. news
	Late fringe:  11 p.m. - 1 a.m.
	Weekend Day:  6 a.m. to 7 p.m.

Marketlo:  Name of media market in which ad was aired

Stncalls:  Call letters of television station on which ad was
aired

Affil:  Network affiliation of television station on which
ad was aired

Sptime:  Time of day at which ad was aired

Cost:  Estimated cost of ad

Showname:  Name of show during which ad was aired

Spotdate:  Date on which ad was aired

Sponsor:
	
Presidential
1	Bush
2	Gore
3	RNC-pres
4	DNC-pres
5	IG-Bush
6	IG-Gore
7	Bush/RNC
8	Gore/DNC
9	Buchanan
10	Browne (Libertarian)
11	Nader
12	Other candidate
13	GOP-unclear
14	Dem-unclear
Presidential Primary
15	GOP - primary
16	Dem - primary
17	IG-GOP primary
18	IG-Dem primary
	
Senate	
21	GOP-candidate
22	Dem-candidate
23	RNC
24	DNC
25	IG-GOP
26	IG-Dem
27	Independent/3rd party candidate
28	GOP-unclear
29	Dem-unclear
Senate Primary
30	GOP primary
31	Dem primary
32	IG-GOP primary
33	IG-Dem primary
	
House	
41	GOP-candidate
42	Dem-candidate
43	RNC
44	DNC
45	IG-GOP
46	IG-Dem
47	Independent/3rd party candidate
48	GOP-unclear
49	Dem-unclear
House Primary
50	GOP primary
51	Dem primary
52	IG-GOP primary
53	IG-Dem primary
	
Governors
61	GOP-candidate
62	Dem-candidate
63	RNC
64	DNC
65	IG-GOP
66	IG-Dem
67	Independent/3rd party candidate
68	GOP-unclear
69	Dem-unclear
Gubernatorial Primary
70	GOP primary
71	Dem primary
72	IG-GOP primary
73	IG-Dem primary
	
Other 	
80	Issue Advocacy
81	Regarding a proposition/other matter
82	Other primary candidacy
84	Vote Republican
85	Vote Democratic
99	Other type of candidacy

Adcode:  Unique numerical identifier given each ad

Statcode:
("AL" =1)
("AK" = 2)
("AZ" =4)
("AR" =5)
("CA" =6)
("CO"=8)
("CT" =9)
("DE" =10)
("DC" =11)
("FL" =12)
("GA" =13)
("HI"=15)
("ID" = 16)
("IL" =17)
("IN" =18)
("IA"=19)
("KS" =20)
("KY" =21)
("LA"=22)
("ME" =23)
("MD" =24)
("MA" = 25)
("MI" = 26)
("MN" =27)
("MS"=28)
("MO" = 29)
("NE" = 31)
("NV" =32)
("NH" = 33)
("NJ" = 34)
("NM" = 35)
("NY" =36)
("NC" = 37)
("ND"=38)
("OH" =39)
("OK" =40)
("OR"=41)
("PA" =42)
("RI" =44)
("SC" =45)
("SD"=46)
("TN"=47)
("TX" =48)
("UT" =49)
("VT"=50)
("VA" =51)
("WA" =53)
("WV" =54)
("WI" =55)
("US" =99)

Statdist:  Unique numerical identifier of each congressional
district.  (Equals statcode*100)+q3

q3.  Race Number
1-55 'House District'
60 'Senate'
70 'Governor'
80 'Issue Advocacy'
95 'Presidential'
99 'Other'

q4.  Is the ad aired for a general election or a primary
election?
1 'Primary'
2 'General'

q5.  What is the party of the favored candidate?
1 'Democrat'	
2 'Republican'	
3 'Other'

q6.  What is the seat's incumbent status?
1 'Open Seat'
2 'Republican Seat'	
3 'Democratic Seat'	
4 'Other (Independent Seat)'

q7.  Does the ad direct the viewer to take any action (as
opposed to merely providing information)?
0 'No'
1 'Yes'

q8.  (If yes to #7) What is that action?
0 'Not applicable'
1 'Other'	
2 'To vote for someone'	
3 'To support someone'	
4 'To elect or re-elect someone'	
5 'To write, call, or tell someone to do something'	
6 'To reject someone'	
7 'To urge action or attention to a particular matter'	
8 'To defeat someone'	
9 'To send a message or call someone to express yourself'	
10 'To vote against someone'	
11 'Other magic word'

q9.  (If an ad asks people to contact a public official)
Does it provide a specific bill number to discuss or urge
action on?
0 'No'
1 'Yes'
2 'Unclear/Unsure'

q10.  (If an ad asks people to contact a public official)
Does it provide a phone number or address to help them do
so?
1 'Toll number listed'	
2 'No'	
3 'Toll-free telephone number listed'	
4 'Address listed'	

q11.  In your opinion, is the purpose of the ad to provide
information about or urge action on a bill or issue, or is
it to generate support or opposition for a particular
candidate?
1 'Generate support or opposition for candidate'	
2 'Provide information or urge action'	
3 'Unsure/unclear'	

q12.  Is the favored candidate�
0 'Not applicable'
1 'Mentioned'	
2 'Pictured in the ad'
3 'Not identified at all'	
4 'Both mentioned and pictured in the ad'

q13.  Is the favored candidate's opponent�.
0 'Not applicable'
1 'Not identified at all'	
2 'Both mentioned and pictured in the ad'	
3 'Pictured in the ad'	
4 'Mentioned by name in the text of an ad'	

q14.  In your judgment, is the primary purpose of the ad to
promote a specific candidate ("In his distinguished career,
Senator Jones has brought millions of dollars home.  We need
Senator Jones"), to attack a candidate ("In his long years
in Washington, Senator Jones has raised your taxes over and
over.  We can't afford 6 more years of Jones.") or to
contrast the candidates ("While Senator Jones has been
raising your taxes, Representative Smith has been cutting
them.")?
0 'Not applicable'
1 'Attack'	
2 'Contrast'	
3 'Promote'	
4 'Unsure or Unclear'

q15.  Does the favored candidate appear on screen narrating
his or her ad?
0 'No'
1'Yes'
9 'Not applicable'

q16.  Is the office at stake mentioned in the ad?
0 'Not applicable'
1 'Yes  referred to in text of the ad'	
2 'No'	
3 'Yes  written in one of the visual frames of the ad'	
4 'Yes  referred to in both the text and visuals of the ad'

q17.  Is an opponent's commercial mentioned or shown on
screen?
0 'Not applicable'
1 'Yes opponents commercial is shown on screen'	
2 'Yes opponents commercial is referred to in text and
screen'	
3 'No'	
4 'Yes opponents commercial is referred to in text'	
	
q18.  Does the ad use any of the following adjective to
characterize the favored candidate?
0 'Not applicable'
1 'Common sense leadership'	
2 'Independent'	
3 'Innovative'	
4 'Self made'	
5 'Caring or Compassionate'	
6 'Bold'	
7 'Principled'	
8 'Tough or a fighter'	
9 'Proven Tested Experienced'	
10 'Values (shares them, has American ones�)'	
11 'No adjectives or descriptions of candidates'	
12 'Protector'	
30 'Other'	
13 'Moderate middle of the road mainstream'	
14 'Conservative'	
15 'Fiscally conservative'	
16 'Hard working'	
17 'Friend of Clinton'	
18 'Committed'	
19 'Visionary'	
20 'Reformer'	
21 'Competent and Knows how to get things done'	
22 'Honest'	
23 'Family man'	

q19.  Second mention (same as #18)

q20.  Does the ad use any of the following adjectives to
characterize the opposing candidate? (first mention)
0 'Not applicable'
1 'Dishonest corrupt'	
2 'Dangerous'	
3 'Friend of Pat Robertson religious right'	
4 'Reckless'	
5 'Too risky'	
6 'Turncoat'	
7 'Incompetent'	
8 'Taxing (or some version of liking taxes)'	
9 'Hypocrite'	
10 'Extremist or radical'	
11 'Career Politician'	
12 'Heartless (may be used in reference to Social Security)'	
13 'Friend of Newt Gingrich'	
14 'Negative'	
15 'Liberal'	
16 'Reactionary or right-wing'	
17 'Friend of special interests'	
18 'No adjectives or description'	
19 'Friend of Clinton'	
30 'Other'	

q21.  Second mention (same as #20)

q22.  Does the ad mention the party label of the favored
candidate or the opponent?
0 'Not applicable'
1 'No'	
2 'Yes  opposing candidates party'	
3 'Yes  both candidates party affiliations are mentioned'	
4 'Yes  favored candidates party'

q23.  Does the ad use technology to distort (i.e. morph)
the opposing candidate's face?
0 'No'
1'Yes'
9 'Not applicable'

q24.  Is the ad funny or is it intended to be humorous?
0 'No'
1'Yes'

q25.  Does the ad refer to newspaper stories or editorials?
1 'Yes  in both the text and visuals'	
2 'No'	
3 'Yes  in the visuals of the ad'	
4 'Yes  in the text part of the text of the ad'	
	
q26.  Does the ad cite supporting sources (including in
footnotes) to bolster various claims?
1 'Yes  in the visuals of the ad'	
2 'No'	
3 'Yes  in the text part of the text of the ad'	

q27.  In your judgment, is the primary focus of this ad on
the personal characteristics of either candidate or on
policy matters?
1 'Policy matters'	
2 'Personal characteristics'	
3 'Both'	
4 'Neither'	

q28.  Does the ad feature a celebrity or a politician
endorsing the candidate?
1 'Celebrity'	
2 'Politician'
3 'Neither'	

q29.  Is the ad in Spanish?
0 'No'
1'Yes'

q30.  Is the ad directly targeted to appeal to a racial
minority?
0 'No'
1'Yes'

q31.  Are the people in the ad racially diverse?
0 'No'
1'Yes'

q32-35:  campaign themes.
1  'Background'	
10 'Taxes'	
11 'Deficit/ surplus/ budget/ debt'	
12 'Government Spending'	
13 'Minimum wage'	
14 'Farming (e.g. friend of)'	
15 'Business (e.g. friend of)'	
16 'Employment/ jobs'	
17 'Poverty'	
18 'Trade/NAFTA'	
19 'Other economic reference'	
2  'Political record'	
20 'Abortion'	
21 'Homosexuality'	
22 'Moral values'	
23 'Tobacco'	
24 'Affirmative action'	
25 'Gambling'	
26 'Assisted Suicide'	
27 'Gun control'	
28 'Other reference to social issues'	
3  'Attendance record'
30 'Crime'	
31 'Drugs'	
32 'Death Penalty'	
33 'Other reference to law and order'	
4  'Ideology'	
40 'Education'	
41 'Lottery for education'	
42 'Child care'	
43 'Other child-related issues'	
5  'Personal values'	
50 'Defense'	
51 'Missile defense/Star wars'	
52 'Veterans'	
53 'Foreign policy'	
54 'Bosnia'	
55 'China'	
59 'Other defense/foreign policy issues'	
6  'Honesty/Integrity'	
60 'Clinton'	
61 'Ken Starr'	
62 'Whitewater'	
63 'Impeachment'	
64 'Sexual harassment/Paula Jones'	
7  'Special Interests'	
70 'Environment'	
71 'Immigration'	
72 'Health care'	
73 'Social security'	
74 'Medicare'	
75 'Welfare'	
76 'Civil rights/race relations'	
77 'Campaign finance reform'	
78 'Government ethics'	
95 'Other'	
99 'None'	
